Linking policy and outcomes: A simulation model of poverty incidence

Many states are striving to meet public demand for accountability by 'benchmarking'-setting social goals for the state and tracking progress in meeting the goals. However, states are finding it difficult to set realistic targets and to assess the impacts of policy on achievement of the targets without a framework that models the relationships among policy targets, policy actions and social and economic forces outside the control of policymakers. This paper develops a dynamic simulation model of one 'benchmark' (poverty incidence) in Oregon, linking transitions into and out of poverty to various events (increased earnings, or having a child as a teenager, for example), and linking these events to policy. The simulation results suggest that, with current policies, Oregon will come close to achieving its poverty benchmark target of 11 percent by the year 2000 if economic conditions remain favorable. The model is used to examine the impact on poverty incidence of three policy strategies: reducing high school dropout and teen pregnancy rates, increasing the effectiveness of social support programs to JOBS participants, and boosting job growth. The simulation results suggest that when assessing the state's performance or 'grading' the observed trend in the poverty benchmark, policymakers should take into account the performance of the state (and national) economy. The impact of policy efforts to reduce poverty is limited because many poverty spells are caused largely by events not affected by current state policies.